Millimeter-Wave (mmWave) Communications Printed Edition of the Special Issue Published in Electronics www.mdpi.com/journal/electronics Manuel García Sanche z Edited by Millimeter-Wave (mmWave) Communications Millimeter-Wave (mmWave) Communications Special Issue Editor Manuel Garc ́ ıa Sanchez MDPI • Basel • Beijing • Wuhan • Barcelona • Belgrade Special Issue Editor Manuel Garc ́ ıa Sanchez Signal Theory and Communications Department, University of Vigo Spain Editorial Office MDPI St. Alban-Anlage 66 4052 Basel, Switzerland This is a reprint of articles from the Special Issue published online in the open access journal Electronics (ISSN 2079-9292) from 2019 to 2020 (available at: https://www.mdpi.com/journal/electronics/ special issues/mmwave commun). For citation purposes, cite each article independently as indicated on the article page online and as indicated below: LastName, A.A.; LastName, B.B.; LastName, C.C. Article Title. Journal Name Year , Article Number , Page Range. ISBN 978-3-03928-430-6 (Pbk) ISBN 978-3-03928-431-3 (PDF) c © 2020 by the authors. Articles in this book are Open Access and distributed under the Creative Commons Attribution (CC BY) license, which allows users to download, copy and build upon published articles, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications. The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons license CC BY-NC-ND. Contents About the Special Issue Editor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii Preface to ”Millimeter-Wave (mmWave) Communications” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Manuel Garc ́ ıa S ́ anchez Millimeter-Wave Communications Reprinted from: Electronics 2020 , 9 , 251, doi:10.3390/electronics9020251 . . . . . . . . . . . . . . . 1 Lorenzo Rubio, Rafael P. Torres, Vicent M. Rodrigo Pe ̃ narrocha, Jes ́ us R. P ́ erez, Herman Fern ́ andez, Jose-Maria Molina-Garcia-Pardo and Juan Reig Contribution to the Channel Path Loss and Time-Dispersion Characterization in an Office Environment at 26 GHz Reprinted from: Electronics 2019 , 8 , 1261, doi:10.3390/electronics8111261 . . . . . . . . . . . . . . 4 Miguel Riob ́ o, Rob Hofman, I ̃ nigo Cui ̃ nas, Manuel Garc ́ ıa S ́ anchez and Jo Verhaevert Wideband Performance Comparison between the 40 GHz and 60 GHz Frequency Bands for Indoor Radio Channels Reprinted from: Electronics 2019 , 8 , 1234, doi:10.3390/electronics8111234 . . . . . . . . . . . . . . 18 Benxiao Cai, Lingling Sun and Yuchao Lei 3D Printing Using a 60 GHz Millimeter Wave Segmented Parabolic Reflective Curved Antenna Reprinted from: Electronics 2019 , 8 , 203, doi:10.3390/electronics8020203 . . . . . . . . . . . . . . . 41 Andrea Massaccesi, Gianluca Dassano and Paola Pirinoli Beam Scanning Capabilities of a 3D-Printed Perforated Dielectric Transmitarray Reprinted from: Electronics 2019 , 8 , 379, doi:10.3390/electronics8040379 . . . . . . . . . . . . . . . 50 Sara Salem Hesari and Jens Bornemann Design of a SIW Variable Phase Shifter for Beam Steering Antenna Systems Reprinted from: Electronics 2019 , 8 , 1013, doi:10.3390/electronics8091013 . . . . . . . . . . . . . . 63 Heng Zhu, Wei Chen, Jianhua Huang, Zhiyu Wang and Faxin Yu A High-Efficiency K-band MMIC Linear Amplifier Using Diode Compensation Reprinted from: Electronics 2019 , 8 , 487, doi:10.3390/electronics8050487 . . . . . . . . . . . . . . . 77 Jihoon Doo, Woojin Park, Wonseok Choe and Jinho Jeong Design of Broadband W-Band Waveguide Package and Application to Low Noise Amplifier Module Reprinted from: Electronics 2019 , 8 , 523, doi:10.3390/electronics8050523 . . . . . . . . . . . . . . . 89 Xiaofan Yang, Xiaoming Liu, Shuo Yu, Lu Gan, Jun Zhou and Yonghu Zeng Permittivity of Undoped Silicon in the Millimeter Wave Range Reprinted from: Electronics 2019 , 8 , 886, doi:10.3390/electronics8080886 . . . . . . . . . . . . . . . 98 Daniel Castanheira, Sara Teodoro, Ricardo Sim ̃ oes, Ad ̃ ao Silva and Atilio Gameiro Multi-User Linear Equalizer and Precoder Scheme for Hybrid Sub-Connected Wideband Systems Reprinted from: Electronics 2019 , 8 , 436, doi:10.3390/electronics8040436 . . . . . . . . . . . . . . . 110 v Adel Aldalbahi Multi-Backup Beams for Instantaneous Link Recovery in mmWave Communications Reprinted from: Electronics 2019 , 8 , 1145, doi:10.3390/electronics8101145 . . . . . . . . . . . . . . 126 Enass Hriba and Matthew C. Valenti Correlated Blocking in mmWave Cellular Networks: Macrodiversity, Outage, and Interference Reprinted from: Electronics 2019 , 8 , 1187, doi:10.3390/electronics8101187 . . . . . . . . . . . . . . 137 Luis Duarte, Rodolfo Gomes, Carlos Ribeiro and Rafael F. S. Caldeirinha A Software-Defined Radio for Future Wireless Communication Systems at 60 GHz Reprinted from: Electronics 2019 , 8 , 1490, doi:10.3390/electronics8121490 . . . . . . . . . . . . . . 154 vi About the Special Issue Editor Manuel Garc ́ ıa S ́ anchez received Telecommunication Engineering degree from the Universidade de Santiago de Compostela, Santiago de Compostela, Spain, in 1990, and a Ph.D. in Telecommunication Engineering from the Universidade de Vigo, Spain, in 1996. He is currently a professor at the Department of Signal Theory and Communications, Universidade de Vigo, Spain. He was head of the department from 2004 to 2010. His research interests include radio systems, indoor and outdoor radio channels, channel sounding and modeling for narrowband and wideband applications, interference detection and analysis, design of impairment mitigation techniques, and radio systems design. vii electronics Editorial Millimeter-Wave Communications Manuel Garc í a S á nchez atlanTTic Research Center, Signal Theory and Communications Department, University of Vigo, 36310 Vigo, Spain; manuel.garciasanchez@uvigo.es Received: 24 January 2020; Accepted: 29 January 2020; Published: 3 February 2020 1. Introduction For the last few decades, the millimeter wave (mmWave) frequency band (30–300 GHz) has been seen as a serious candidate to host very high data rate communications. First used for high capacity radio links, then for broadband indoor wireless networks, the interest in this frequency band has boosted, as it is proposed to accommodate future 5G mobile communication systems. The large bandwidth available in this frequency band will enable a number of new use cases for 5G. In addition, due to the large propagation attenuation, this frequency band may present some additional advantages regarding frequency reuse and communication security. On the other hand, however, a number of issues have to be addressed to make 5G mmWave communications viable: radio channel measurement, modeling, and estimation; antenna design and antenna measurement; beamforming and energy e ffi ciency; commercial hardware design and development; multiple-input multiple-output (MIMO) and massive MIMO (m-MIMO) techniques; multi-cell cooperation; network planning and interference; system performance assessment and optimization; and finally, the study of new case uses and applications. 2. Contributions in This Special Issue Each of the twelve papers collected in this Special Issue contributes to a solution to one or more of the challenges described in the introduction. Regarding radio wave propagation, Rubio et al. [ 1 ] provide an experimental characterization of the path loss and time-dispersion of an in o ffi ce radio channel at 26 GHz, while in a study by Riob ó and colleagues [ 2 ], wideband results at 40 GHz and 60 GHz frequency bands are also provided for indoor environments. Two other papers [ 3 , 4 ] deal with the design and assessment of di ff erent kinds of antennas manufactured using three-dimensional (3D) printing. In one case, a 60 GHz segmented parabolic reflective curved antenna, with a gain of 20 dBi at 64 GHz, is presented by Cai, Sun, and Lei [ 3 ], while Massaccesi, Dassano, and Pirinoli [ 4 ] design a perforated dielectric transmitarray and analyze its beam scanning capabilities. Also related to beam steering antennas, Salem Hesari and Bornemann [ 5 ] describe the design, fabrication, and assessment of a substrate integrated waveguide variable phase shifter that may steer the radiation pattern of the antenna by ± 25º. Another challenge for mmWave communication system is the design of amplifiers. Two di ff erent kinds of mmWave amplifiers are presented in two contributions. The design of a high-e ffi ciency K-band MMIC linear amplifier using diode compensation is presented by Zhu and co-workers [ 6 ] together with its measured performance, while in a study by Doo and colleagues [ 7 ] a broadband mmWave waveguide package, which covers the entire W-band (75–110 GHz), is presented and applied to build a low noise amplifier module. This module measures gains greater than 14.9 dB from 75 GHz to 105 GHz (12.9 dB at the entire W-band) and noise figures less than 4.4 dB from 93.5 GHz to 94.5 GHz. For a proper design of the electronic systems at mmWave frequencies a good empirical characterization of the dielectric properties of the substrate material is of capital importance. In a study by Yang et al. [ 8 ] a description of the dielectric measurement of undoped silicon in the E-band (60–90 GHz) using a free-space quasi-optical system is provided. Electronics 2020 , 9 , 251; doi:10.3390 / electronics9020251 www.mdpi.com / journal / electronics 1 Electronics 2020 , 9 , 251 Precoding is other key technology that will enable 5G mmWave communications. Castanheira and co-workers [ 9 ] present a sub-connected hybrid analog / digital multi-user linear equalizer combined with an analog precoder to e ffi ciently remove the multi-user interference. As propagation environment imposes several restrictions to radio wave propagation at mmWave frequencies and strong link blockage may occur, any technique to facilitate link recovery constitutes a significant contribution. A multibeam technique to speed up link recovery is presented by Aldalbahi [ 10 ]. Hriba and Valenti [ 11 ] discuss another way to mitigate link blocking by using macrodiversity techniques, however the performance of macrodiversity can be reduced if correlated blocking occurs in links to di ff erent base stations. Finally, Duarte and colleagues [ 12 ] present a complete end-to-end 5G mmWave testbed fully reconfigurable based on a FPGA architecture. 3. Future Trends The development of millimeter-wave communication systems has just started. Despite the recent developments to cope with the multiple challenges that researchers should solve, there is still a lot of work to be done. During the next years, we hope to assist in the exponential growth of contributions to this field: mm-Wave communications will lead us to full development of 5G case studies and beyond. Acknowledgments: I would like to thank, in the first place, all the authors who decided to send they research contributions to this Special Issue. I also would like to thank all the reviewers for their comments that helped to improve the papers in this issue. Finally, I would like to thank MDPI editorial board and sta ff for the opportunity to guest-edit this Special Issue. Conflicts of Interest: The author declares no conflicts of interest. References 1. Rubio, L.; Torres, R.; Rodrigo Peñarrocha, V.; P é rez, J.; Fern á ndez, H.; Molina-Garcia-Pardo, J.M.; Reig, J. Contribution to the Channel Path Loss and Time-Dispersion Characterization in an O ffi ce Environment at 26 GHz. Electronics 2019 , 8 , 1261. [CrossRef] 2. Riob ó , M.; Hofman, R.; Cuiñas, I.; Garc í a S á nchez, M.; Verhaevert, J. Wideband Performance Comparison between the 40 GHz and 60 GHz Frequency Bands for Indoor Radio Channels. Electronics 2019 , 8 , 1234. [CrossRef] 3. Cai, B.; Sun, L.; Lei, Y. 3D Printing Using a 60 GHz Millimeter Wave Segmented Parabolic Reflective Curved Antenna. Electronics 2019 , 8 , 203. [CrossRef] 4. Massaccesi, A.; Dassano, G.; Pirinoli, P. Beam Scanning Capabilities of a 3D-Printed Perforated Dielectric Transmitarray. Electronics 2019 , 8 , 379. [CrossRef] 5. Salem Hesari, S.; Bornemann, J. Design of a SIW Variable Phase Shifter for Beam Steering Antenna Systems. Electronics 2019 , 8 , 1013. [CrossRef] 6. Zhu, H.; Chen, W.I.; Huang, J.; Wang, Z.; Yu, F. A High-E ffi ciency K-band MMIC Linear Amplifier Using Diode Compensation. Electronics 2019 , 8 , 487. [CrossRef] 7. Doo, J.; Park, W.; Choe, W.; Jeong, J. Design of Broadband W-Band Waveguide Package and Application to Low Noise Amplifier Module. Electronics 2019 , 8 , 523. [CrossRef] 8. Yang, X.; Liu, X.; Yu, S.; Gan, L.; Zhou, J.; Zeng, Y. Permittivity of Undoped Silicon in the Millimeter Wave Range. Electronics 2019 , 8 , 886. [CrossRef] 9. Castanheira, D.; Teodoro, S.; Sim õ es, R.; Silva, A.; Gameiro, A. Multi-User Linear Equalizer and Precoder Scheme for Hybrid Sub-Connected Wideband Systems. Electronics 2019 , 8 , 436. [CrossRef] 10. Aldalbahi, A. Multi-Backup Beams for Instantaneous Link Recovery in mmWave Communications. Electronics 2019 , 8 , 1145. [CrossRef] 2 Electronics 2020 , 9 , 251 11. Hriba, E.; Valenti, M.C. Correlated Blocking in mmWave Cellular Networks: Macrodiversity, Outage, and Interference. Electronics 2019 , 8 , 1187. [CrossRef] 12. Duarte, L.; Gomes, R.; Ribeiro, C.; Caldeirinha, R. A Software-Defined Radio for Future Wireless Communication Systems at 60 GHz. Electronics 2019 , 8 , 1490. [CrossRef] © 2020 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http: // creativecommons.org / licenses / by / 4.0 / ). 3 electronics Article Contribution to the Channel Path Loss and Time-Dispersion Characterization in an Office Environment at 26 GHz Lorenzo Rubio 1, *, Rafael P. Torres 2 , Vicent M. Rodrigo Peñarrocha 1 , Jesús R. Pérez 2 , Herman Fernández 3 , Jose-Maria Molina-Garcia-Pardo 4 and Juan Reig 1 1 iTEAM Research Institute, Universitat Politècnica de València, 46022 Valencia, Spain; vrodrigo@dcom.upv.es (V.M.R.P.); jreigp@dcom.upv.es (J.R.) 2 Departamento de Ingeniería de Comunicaciones, Universidad de Cantabria, 39005 Santander, Spain; rafael.torres@unican.es (R.P.T.); jesusramon.perez@unican.es (J.R.P.) 3 Escuela de Ingeniería Electrónica, Universidad Pedagógica y Tecnológica de Colombia, Sogamoso 152211, Colombia; herman.fernandez@uptc.edu.co 4 Departamento de Tecnologías de la Información y las Comunicaciones, Universidad Politécnica de Cartagena, Cartagena, 30202 Murcia, Spain; josemaria.molina@upct.es * Correspondence: lrubio@dcom.upv.es; Tel.: +34-963879739 Received: 7 October 2019; Accepted: 28 October 2019; Published: 1 November 2019 Abstract: In this paper, path loss and time-dispersion results of the propagation channel in a typical office environment are reported. The results were derived from a channel measurement campaign carried out at 26 GHz in line-of-sight (LOS) and obstructed-LOS (OLOS) conditions. The parameters of both the floating-intercept (FI) and close-in (CI) free space reference distance path loss models were derived using the minimum-mean-squared-error (MMSE). The time-dispersion characteristics of the propagation channel were analyzed through the root-mean-squared (rms) delay-spread and the coherence bandwidth. The results reported here provide better knowledge of the propagation channel features and can be also used to design and evaluate the performance of the next fifth-generation (5G) networks in indoor office environments at the potential 26 GHz frequency band. Keywords: 5G; mmWave; path loss; time-dispersion; delay-spread; coherence bandwidth; channel measurements 1. Introduction Some of the main objectives proposed in the deployment of the future fifth-generation (5G) systems are the increase in the data rate and capacity, greater than 100 Mbps, with peak data rates up to 10 Gbps, ultra-reliable and low-latency communications, and communications in high user density scenarios [ 1 , 2 ]. The first 5G deployments, at least at the European level, will be carried out in the harmonized frequency bands below 1 GHz, in particular the 700 MHz band, corresponding to the second digital dividend, together with the 3.4–3.8 GHz frequency band [ 3 ]. However, the high transmission rates expected in 5G can only be achieved using the spectrum at frequencies above 24 GHz, where it is possible to use bandwidths of the order of hundreds of megahertz [ 2 ]. At the last World Radiocommunication Conference (WRC) of the International Telecommunication Union (ITU), held in 2015 (WRC-15), the potential bands to locate future 5G systems, on a primary basis, above 24 GHz are: 24.25–27.5 GHz, 31.8–33.4 GHz, and 37–40.5 GHz, [4]. Although the final decision will be conditioned in part by the industry, the potential for global harmonization, and research works, there is some consensus to deploy the 5G systems in the 26 GHz frequency band. In fact, the Radio Spectrum Policy Group (RSPG) has recommended the 24.25–27.5 GHz band for the 5G deployments in Europe. Electronics 2019 , 8 , 1261; doi:10.3390/electronics8111261 www.mdpi.com/journal/electronics 4 Electronics 2019 , 8 , 1261 Many measurement campaigns in both indoor and outdoor environments have been conducted at some typical millimeter wave (mmWave) bands (although the mmWave band strictly corresponds to frequencies between 30 and 300 GHz, it is common in the literature to also consider frequencies above 10 GHz), in particular at 11, 15, 28, 38, 60, and 73 GHz [ 5 – 9 ]. Nevertheless, little attention has been devoted to the 26 GHz band. Although the propagation characteristics measured at 28 GHz could be extrapolated to the 26 GHz frequency band, specific channel measurements are necessary for better knowledge of the propagation channel. In this sense, a contribution to the path loss and time-dispersion characterization, in terms of the delay-spread and the coherence bandwidth, is performed in this paper. The study is based on a channel measurement campaign at 26 GHz carried out in an indoor office environment. The measurements were collected under line-of-sight (LOS) and obstructed-LOS (OLOS) conditions with a channel sounder implemented in the frequency domain using a vector network analyzer (VNA) and a broadband radio over fiber (RoF) link to increase the dynamic range in the measurement and allowing us to use omnidirectional antennas at the transmitter (Tx) and the receiver (Rx). The remainder of the paper is organized as follows. Section 2 describes the propagation environment, measurement setup, and procedure. In Section 3, path loss and time-dispersion results are presented and discussed. Finally, conclusions are drawn in Section 4. 2. Channel Measurements 2.1. Propagation Environment The channel measurements were carried out in an office environment, characterized by the presence of desks, with PC monitors, chairs, and some steel storage cabinets, among other typical objects in these environments. The office was in a modern building construction with large exterior glass windows, where the ceiling and the floor were built of reinforced concrete over steel plates with wood and plasterboard paneled walls. Figure 1 shows a panoramic view of the office. The propagation environment consisted of a 9.68 m by 6.93 m room with a height of 2.63 m. Figure 1. Panoramic view of the propagation environment. 2.2. Measurement Setup and Procedure A channel sounder was implemented in the frequency domain to measure the complex channel transfer function (CTF), denoted as H ( f ) . The channel sounder was based on the Keysight N5227A VNA. The QOM-SL-0.8-40-K-SG-L ultra-wideband antennas, developed by Steatite Ltd company, were used at the Tx and Rx sides. These antennas operate from 800 MHz to 40 GHz, have an omnidirectional radiation pattern in the azimuth plane (horizontal plane), and linear polarization. Figure 2 shows the three-dimensional (3D) radiation pattern of the antennas measured in our anechoic chamber. The 3 dB beamwidth of the antennas in the elevation plane, known as half power beamwidth (HPBW), was in the order of 35 ◦ at 26 GHz. 5 Electronics 2019 , 8 , 1261 Figure 2. 3D radiation pattern of the antenna at 26 GHz. The Tx antenna was connected to the VNA through an amplified broadband RoF link, the Optica OTS-2 model developed by Emcore (from 1 to 40 GHz, with 35 dB of gain). This RoF link avoided the high losses of cables at mmWave frequencies, thus increasing the dynamic range in the measurement and allowing us to use omnidirectional antennas due to the fact that significant features of the propagation channel, such as time-dispersion, could be affected by the use of directional antennas [ 10 ]. The Rx antenna was located in a XY positioning system, implementing a 12 × 12 uniform rectangular array (URA). The inter-element separation of the URA was 3.04 mm. This separation was about λ / 4 at 26 GHz, covering the Rx antenna with a local area around ( 11 / 4 ) 2 λ 2 in order to take into account small-scale fading effects. Both the VNA and the XY positioning system were controlled by a personal computer. The S 21 ( f ) scattering parameter, equivalent to H ( f ) [ 11 ], was measured from 25 to 27 GHz, i.e., a channel bandwidth (SPAN in the VNA) of 2 GHz was employed with 26 GHz as a central frequency. Notice that a VNA measures the scattering parameters of a device under test (DUT). In this case, the DUT was the wireless channel, where the S 21 ( f ) scattering parameter was the CTF at the frequency that was used to excite the channel. By having the excitation signal sweep through the frequency band of interest, i.e., the SPAN, a sampled version of the CTF was measured. The radiofrequency signal level at the VNA was − 17 dBm to not saturate the amplifier at the input of the electro-optical converter in the RoF link. Before the measurements, the channel sounder was calibrated carefully. A response calibration process was performed, moving the time reference points from the VNA port to the calibration points. Thus, the measured CTF took into account the joint response of the propagation channel and the Tx and Rx antennas, also known in the literature as the radio channel [12]. A schematic diagram of the channel sounder is shown in Figure 3. A total of N f = 1091 frequency points was measured over the 2 GHz bandwidth. Thus, the frequency resolution was about Δ f ≈ 1.83 MHz (2 GHz/ N f ), which corresponded to a maximum unambiguous excess delay estimated as 1 / Δ f of 546 ns. This maximum unambiguous excess delay was equivalent to a maximum observable distance calculated as c 0 / Δ f , with c 0 the speed of light, of about 164 m. Notice that the maximum observable distance was larger than the office dimensions, ensuring that all multipath contributions were captured. The bandwidth of the intermediate frequency (IF) filter at the VNA, denoted by B IF , was set to 100 Hz. This value of B IF was a trade-off between acquisition time and dynamic range in the measurement. Thus, low values of B IF reduced the noise floor, increasing the dynamic range in the measurement. Nevertheless, low values of B IF increased the acquisition time. As a reference, in [ 7 ], the authors used 500 Hz in indoor office channel measurements at mmWave frequencies. Table 1 summarizes the measurement system parameters. 6 Electronics 2019 , 8 , 1261 Table 1. Measurement system parameters. Parameter Value VNA output power − 17 dBm VNA center frequency 26 GHz VNA SPAN(Bandwidth) 2 GHz VNA IF Bandwidth ( B IF ) 100 Hz Frequency points per trace 1091 Tx/Rx antenna gain 5.2 dB Tx antenna height 0.90 m Rx antenna height 1.62 m 5DGLRFKDQQHO *+] 3RUW 2XW 3RUW ,Q 9HFWRU1HWZRUN$QDO\]HU 7[ DQWHQQD 5[ DQWHQQD 3& ;<SRVLWLRQLQJ V\VWHP 85$ 7ULSRG 5R) OLQN 6 I + I { Figure 3. Schematic diagram of the channel sounder. During the measurements, the Tx antenna was located manually in different locations of the office, imitating the position of user equipment (UE). The Rx subsystem remained fixed in the same position, close to the wall, imitating an access point (AP) that served the users inside the office. The Rx antenna height was 1.62 m with respect to the floor. With the VNA configuration parameters, i.e., N f and B IF , the acquisition time to capture the CTF in the 144 (12 × 12) positions of the Rx antenna in the URA was about 2 h. To guarantee stationary channel conditions (due to the frequency sweep time to measure the S 21 ( f ) scattering parameter, the acquisition of the channel measurements required stationary channel conditions), the measurements were collected at night, thus avoiding the presence of people, not only in the measurement room, but also in adjoining areas. Figure 4 shows the top view of the propagation environment, together with the Rx-URA position and the Tx antenna locations. The channel measurements were performed in LOS and OLOS conditions, defining two scenarios: • Scenario LOS: The Tx antenna was located at a height of 0.90 m above the floor level, imitating the position of a UE (e.g., a laptop, tablet, or mobile phone) that was on the desk. A total of 10 Tx locations (Tx1–Tx10) was considered in the measurements. Figure 5 (left) shows a view of the Rx-URA and the Tx antenna for the Tx1 position. • Scenario OLOS: The Tx antenna was also located at a height of 0.90 m above the floor and close to the desk, but in OLOS propagation conditions due to the blockage of the direct component by the computer monitors on the desks. The measurements were taken in 4 Tx locations (Tx11-Tx14). Figure 5 (right) shows a view of the Rx-URA and the Tx antenna for the Tx14 position. 7 Electronics 2019 , 8 , 1261 ĞƐŬ ŚĂŝƌ 'ůĂƐƐ KĨĨŝĐĞĐůŽƐĞƚ &ƌŝĚŐĞ ddžϭ ddžϮ ddžϯ ddžκ ddžρ ddžς ddžϳ ddžΘ ddžε ddžϭϬ ddžϭϭ ddžϭϮ ddžϭϯ ddžϭκ ddž K>K^ ddž >K^ Zdž Ͳ hZ ZdžͲhZ WŵŽŶŝƚŽƌ Figure 4. Top view of the propagation environment. The Rx-URA and the Tx antenna locations are indicated. 5[ 7[ /26 Zdž ddžϭκ;K>K^Ϳ Figure 5. View of the Rx antenna and the Tx antenna in positions Tx1 in LOS (left) and Tx14 in OLOS (right). 3. Measurement Results 3.1. Path Loss Results For each position of the URA, the path loss between the Tx and Rx antennas can be derived by averaging the CTF in frequency and taking into account the gain of both antennas in the direct path [13]. Thus, the path loss in logarithmic units, PL , can be derived as (see Appendix A): PL ( d ) = − 10 log 10 ⎛ ⎝ 1 N f N f ∑ n = 1 | H ( f n , d ) | 2 g Tx ( f n ) g Rx ( f n ) M ( f n ) ⎞ ⎠ , (1) where d refers to the separation distance between the Tx antenna and the center of the URA for each Tx location, indicated as Tx-Rx distance; f n is the n th frequency sample; and g Tx ( f n ) and g Rx ( f n ) are the gain of the Tx and Rx antennas, respectively, in the direction defined by the direct path contribution. The term M ( f n ) takes into account the mismatch of the antennas, and it is calculated by: M ( f n ) = ( 1 − | S Tx 11 ( f n ) | 2 )( 1 − | S Rx 11 ( f n ) | 2 ) , (2) 8 Electronics 2019 , 8 , 1261 S Tx 11 ( f n ) and S Rx 11 ( f n ) being the S 11 ( f ) scattering parameter of the Tx and Rx antennas, respectively. The measured path loss (cross marker) for each Rx antenna position in the URA in terms of the Tx-Rx distance is shown in Figure 6. Both LOS and OLOS propagation conditions were considered. It is worth noting that the spread values of the path loss along the URA due to the short-term fading was less than 2.5 dB, being less in LOS than in OLOS conditions. It can be observed that the spread values of the path loss did not exhibit any correlation with the Tx-Rx distance. The mean value of the path loss (square marker) for each Tx location is also depicted in Figure 6. 2 3 4 5 6 7 8 9 Tx-Rx distance (m) 60 65 70 75 80 Path loss(dB) Measured LOS Measured OLOS Mean LOS Mean OLOS FI LOS model CI LOS model FI OLOS model CI OLOS model Figure 6. Path loss in terms of the Tx-Rx distance. Measured data, measured mean values, and estimated values from the CI and FI models, in both LOS and OLOS conditions. The floating-intercept (FI) path loss model has been widely used to describe the behavior of the path loss in terms of the Tx-Rx distance in the microwave frequency band, particularly at the sub-6 GHz band and more recently in mmWave frequencies [ 5 , 14 ], being one of the propagation models adopted in channel standardizations, e.g., the WINNERII Project and 3GPP channel models [ 15 , 16 ]. From the FI model, the path loss is given by: PL FI ( d ) = β + 10 α log 10 ( d ) + χ FI σ , (3) β being the floating-intercept parameter (an offset term); α the path loss exponent, related to both the environment and propagation conditions; and χ FI σ a zero mean Gaussian random variable, in logarithmic units, with standard deviation σ , which describes the large-scale signal fluctuations about the mean path loss over distance, also known in the literature as the shadow factor (SF). The FI model has a mathematical curve fitting approach over the measured path loss set without any physical anchor. On the other hand, the close-in (CI) free space reference distance path loss model is also adopted in many studies related to mmWave propagation [5,8]. In the CI model, the path loss is given by: PL CI ( d ) = FSPL ( f , 1 m ) + 10 n log 10 ( d ) + χ CI σ , (4) where FSPL ( f , 1 m ) = 10 log 10 ( 4 π f / c 0 ) 2 is the free space path loss for a Tx-Rx distance equal to 1 m, with c 0 the speed of light; n is the path loss exponent; and χ CI σ is the SF term. Note that the CI model has certain physical support in the sense that there is an intrinsic frequency dependence of the path loss included in the 1 m FSPL term. Taking into account that FSPL ( f , 1 m ) is equal to 60.74 dB at 26 GHz, (4) can be rewritten as: PL CI ( d ) = 60.74 + 10 n log 10 ( d ) + χ CI σ (5) 9 Electronics 2019 , 8 , 1261 The path loss fitting results for the FI and CI models are also depicted in Figure 6. Both models exhibit a good fit and predict similar path loss values for the Tx-Rx distance considered, particularly in OLOS conditions. It is worth noting that the maximum path loss difference between LOS and OLOS conditions was about 5 dB, increasing with the Tx-Rx distance. Tables 2 and 3 summarize the mean value and their 95% confidence interval of the FI and CI model parameters. These parameters were derived from the measured path loss using the minimum-mean-squared-error (MMSE) approach. Table 2. FI path loss model parameters. β ( β 95% ) α ( α 95% ) σ (dB) LOS 59.29 (58.80–59.79) 1.46 (1.39–1.53) 1.73 OLOS 60.01 (59.46–60.16) 1.88 (1.80–1.95) 0.92 Table 3. CI path loss model parameters. FSPL (1 m) n ( n 95% ) σ (dB) LOS 60.74 dB 1.27 (1.26–1.28) 1.75 OLOS 60.74 dB 1.79 (1.78–1.80) 0.93 For the FI model, α had a mean value equal to 1.46, with 1.39–1.53 the 95% confidence interval, in LOS conditions. In OLOS conditions, α had a mean value equal to 1.88, with 1.80–1.95 the 95% confidence interval. For the CI model, α had a mean value equal to 1.27 and 1.79 for LOS and OLOS conditions, respectively. The 95% confidence intervals were narrower for the CI model. In both models, the SF had a similar value, being lower in OLOS conditions. The values of the path loss exponent derived in this study were lower than the values reported in [ 8 ], where path loss exponents in the order or 2.0 and 2.2 were measured at 28 GHz in LOS and non-LOS (NLOS) conditions, respectively, for the FI model. Nevertheless, higher values have been reported for the CI model, where the path loss exponents equal to 1.45 and 2.18 have been measured in LOS and NLOS conditions, respectively. These differences can be explained because the frequencies are slightly different and, of course, due to both the particular characteristics of the environments and propagation conditions. It is worth noting that in our OLOS measurements, only a few MPCs were blocked by the PC monitors, whereas in NLOS conditions, the Tx and Rx were usually separated by different obstructions, and in many cases, the Tx and Rx were not located in the same room. Despite this, our results were more in line with those published by Rappaport et al. in [ 5 , 17 ] for indoor environments at 28 GHz in LOS conditions, where exponents equal to 1.2 and 1.1 were derived for the FI and CI models, respectively, considering omnidirectional path loss modeling (Rappaport et al. used a sliding correlation channel sounder, synthesizing an omnidirectional path loss model from directional measurements). Furthermore, the SF derived was 1.8 dB in both the FI and CI model, a value very close to that obtained by us for the CI model (1.75 dB). 3.2. Time-Dispersion Results In wideband systems, the multipath propagation causes the arriving signal at the Rx to have a longer duration than the transmitted signal. This effect is the well-known time-dispersion of a wireless channel [ 11 ]. In the frequency domain, the time-dispersion can be interpreted as frequency selectivity, i.e., the CTF varies over the bandwidth of interest. The knowledge of the time-dispersion of a wireless channel is vital in the design of wireless systems, for example adopting efficient equalizer structures or defining the optimal multicarrier separation in digital modulations and diversity schemes. In this section, the time-dispersion of the propagation environment is analyzed in both the delay (time) and frequency domains. 10 Electronics 2019 , 8 , 1261 3.2.1. Root-Mean-Squared Delay-Spread The root-mean-squared (rms) delay-spread, denoted by τ rms , is the most relevant parameter to describe the time-dispersion of a wireless channel in the delay domain. τ rms corresponds to the second-order central moment of the power delay profile (PDP), and it is derived as [18]: τ rms ( d ) = √ √ √ √ ∑ N f n = 1 ( τ n − ̄ τ ( d )) 2 PDP ( τ n , d ) ∑ N f n = 1 PDP ( τ n , d ) , (6) where τ n refers to the n th delay bin in the PDP. Assuming ergodicity [ 11 ], the PDP can be estimated averaging the channel impulse response (CIR), denoted by h ( τ , d ) , over all positions of the Rx antenna in the URA: PDP ( τ , d ) = E m {| h ( τ , d ) | 2 } (7) The CIR is obtained as the inverse Fourier transform of H ( f , d ) . As an example, Figure 7 shows the PDP measured in the Tx1 (LOS condition) and Tx14 (OLOS condition) positions. The differences between LOS and OLOS conditions for low delays can be observed, where the PC monitors blocked the first MPCs, in this case with excess delays around 50 ns. ̄ τ ( d ) = ∑ N f n = 1 τ n PDP ( τ n , d ) ∑ N f n = 1 PDP ( τ n , d ) (8) The cumulative distribution function (CDF) of τ rms for the LOS and OLOS scenarios is shown in Figure 8. A threshold of 30 dB and a Hamming windowing method were considered in the derivation of τ rms Both curves exhibited a similar trend, with a separation of the order of 3 ns around the mean values. The results showed that the time-dispersion was slightly higher in the OLOS scenario. The minimum, mean, maximum, and standard deviation (STD) values of τ rms are summarized in Table 4. For the LOS scenario, τ rms ranged from 11.21 to 21.74 ns, with a mean value equal to 15.88 ns; whereas in the OLOS scenario, the values were higher, with τ rms ranging from 15.13 to 27.87 ns, with a mean value equal to 18.87 ns, 3 ns more than in the LOS scenario. Nevertheless, the STD of τ rms was very similar in both scenarios, in the order of 2 ns. It is worth noting that the values of τ rms derived here were higher than those published in [ 7 ] for an office environment, where values of 8 and 10 ns were reported at 28 and 38 GHz, respectively, in LOS conditions. 0 50 100 150 200 250 300 350 Delay (ns) -140 -130 -120 -110 -100 -90 -80 -70 Average received gain (dB) LOS, Tx1 OLOS, Tx14 Direct component Multipath components Figure 7. PDP measured in Tx1 and Tx14 positions with LOS and OLOS propagation conditions, respectively. 11